Understanding the Effects of Red Tape Costs on the Performance of Firms Exports: The Case of Egypt *

Transcription

1 Understanding the Effects of Red Tape Costs on the Performance of Firms Exports: The Case of Egypt * Rana Hendy Chahir Zaki August, 2013 Abstract This paper tackles the impact of red tape barriers on firms exports. The topic of this paper is crucial in international trade for three main reasons: first, trade barriers- as argued by the WTO- are highly correlated to lengthy, bureaucratic and time consuming trade procedures that do negatively affect firms exports. Second, these barriers are highly persistent and costly in developing countries such as Egypt. In addition, they represent a deadweight loss as they do not generate any rent or revenue. In the present study, we estimate a gravity model using Egyptian firm-level data to examine the impact of these barriers on firms exports. For administrative barriers, we use the Doing Business dataset that has been developed by the World Bank. The preliminary findings show that red tape barriers negatively affect Egyptian firms. This effect seems to be more robust for the intensive margin (the value of exports) rather than the extensive one (number of products and firms per destination). Yet, the impact of administrative barriers to trade on the number of destinations by firm is likely to be significantly negative. Keywords: Gravity, Trade Barriers, Firm-level data, Egypt. JEL codes: F13, F15, F14 1. Introduction * We gratefully acknowledge the General Organization for Export and Import Control (GOEIC), the Ministry of Industry and Foreign Trade in Egypt, for providing us with the firm-level data used in this study. We are also grateful to the World Bank team for their useful comments. Economist, Economic Research Forum (ERF), Cairo, Egypt. rhendy@erf.org.eg Assistant Professor, Faculty of Economics and Political Science, Cairo University, Cairo, Egypt. chahir.zaki@feps.edu.eg

2 Empirical literature has shown that trade costs are important. This is why Anderson and van Wincoop (2004) argue that trade costs are large, even aside from trade policy barriers and even between apparently highly integrated economies.. Such trade costs may be divided in four categories: first, transaction costs related to transport (including distance) and insurance of traded goods; second, costs induced by trade policies associated with tariff and non-tariff barriers (such as quotas, sanitary and phyto-sanitary measures (SPS) and technical barriers to trade (TBT)); third, local distribution costs from foreign producer to final user in the domestic country; and finally costs due to administrative barriers or red tape costs, i.e. the ones associated with trade facilitation. Thus, trade facilitation, that aims at making international trade easier, encompasses a large range of facets that can be summarized in four main points: simplification of trade procedures; harmonization of commercial rules; transparent information and procedures; the recourse to new technologies allowing trade promotion and more secured means of payment. Several main reasons explain why trade facilitation has an increasing policy relevance. First, over the last decades, intra-industry trade and intermediate products trade have significantly increased. These types of trade require quicker and more efficient delivery, especially with a higher interdependency of supply chains. Second, since trade liberalization is not on its own sufficient to ensure the integration of an economy in the globalized world, a particular attention has been turned to understand what are the other impediments that hinder international trade, such as these administrative barriers. Third, administrative procedures are complicated, lengthy and redundant and do have a high cost. Their cost accounts for 2% to 15% of the value of traded goods (OECD, 2002). This is why their elimination is likely to have a highly positive impact on both international trade and welfare. In addition to these reasons, administrative barriers to trade hinder export diversification. This diversification can be observed at both the intensive (the value of exports) and the extensive (number of products and firms per destination). In fact, theory predicts that lowering administrative barriers to trade, e.g., through trade facilitation measures, would make it profitable even for low-productivity firms to become exporters. Trade facilitation cannot only increase domestic productivity within a country, but also promote the entry of new firms into export markets (extensive margin). Furthermore, lower trade costs may also lead to a higher export intensity and stimulate the growth of exports (intensive margin). The reason behind this is explained by Dennis and Ben Shepherd (2011) who argued, that lower fixed costs of export (such as the barriers associated to trade facilitation) expand the range of products that developing countries can export. Fixed costs are perceived as the primary determinants of firm entry into particular overseas

3 product markets. For this reason, they found that a 10% improvement in trade facilitation is associated with product diversity gains of the order of 3%-4%. Moreover, there is evidence that differentiated goods (such as manufactures) have stronger diversification responses to trade facilitation (measured as a uniform proportionate cut in administration costs) than do homogeneous goods (such as agricultural products). The literature on export performance and trade facilitation using firm level data in MENA region is highly scarce since most of the work has been done for either Asian, African or Latin American economies. For Asia countries, Li and Wilson (2009) found that improvement in trade facilitation indicators tend to increase the probability that Small and Medium Enterprises (SMEs) will become exporter as well as their export propensity. As per Latin American economies, Cirrera et al (2011) used a micro dataset that links production, trade and innovation data at the firm level in Brazil and explored the impact of innovation, trade and production dynamics on diversification behavior. In Africa, Yoshino (2008) found that public infrastructure constraints, such as customs delays, seem to have immediate impacts on regional exports in general, implying the relevance of addressing behind-the-border constraints in fostering regional integration. In the same line, Hoekstra (2013) found that trade facilitation can increase African firms probability to participate in international trade. Moreover, lower trade barriers are associated with a higher growth of exports. Shepherd (2012) proved that licensing times do matter for the ability of firms to access imported intermediates. He provided evidence on the fact that clearance times matter for firm-level export performance and that clearance times affect firms choice to export directly or through a third party since longer clearance times make use of a third-party distributor more likely. While numerous studies have highlighted MENA s weak performance in aggregate trade and diversification (Dogruel and Tekce, 2011), little is known about the impact of administrative barriers on firm-level exporter behavior largely because of a lack of data. For this reason, using a recently-available dataset on firm exports in Egypt, this paper fills a knowledge gap in our understanding of how red tape barriers affect exports performance of Egyptian firms. It is worthy to mention that aggregate data hide a lot of heterogeneity and do not allow us to distinguish which firms drive growth and diversification, nor which margins of adjustments matter most. Understanding exactly how the process of trade facilitation affects export growth happens is necessary to identify the drivers of and constraints to export growth in the region and requires microdata. In the present study, we estimate a gravity model using Egyptian firm-level data, for the first time, to examine the impact of these barriers on firms exports. Customs level data on exporting and importing are merged with administrative barriers that come from the Doing Business dataset developed by the World Bank. The preliminary findings show

4 that red tape barriers negatively affect Egyptian firms. This effect seems to be more robust for the intensive margin (the value of exports) rather than the extensive one (number of products and firms per destination). Yet, the impact of administrative barriers to trade on the number of destinations by firm is likely to be significantly negative. In what follows, section 2 presents some stylized facts of the Egyptian exports at the macroeconomic as well as the sectors levels. Section 3 shows how red tape barriers matter for Egyptian exports. Section 4 exhibits the methodology adopted in our study. Section 5 is devoted to the data presentation. In section 6, we present the empirical results. And, section 7 concludes and presents the policy implications of the study. 2. Exports Landscape 2.1. Aggregate export flows Both exports and imports in Egypt experienced significant increases since early 1990s and in a more pronounced way after Figure 1 plots the evolution of exports and imports from 1990 to On the one hand, Figure 1 shows that both exports and imports increase after 2004 are much higher than those before On average, exports increased annually by 5% before 2004 vs. 24% after this date, while imports by 2% and 24% respectively. The same analysis applies for imports. On the other hand, Egypt trade balance has been continuously in deficit throughout the period of the study. Imports exceed exports as a result of the upsurge in the volume of imports that are mainly concentrated in raw materials, investment goods or semi-finished products that are used in the production process. Figure 1: Egypt's Total Exports and Imports of Goods and Services Export Proceeds Services Receipts Import Payments Services Payments Source: Central Bank of Egypt.

5 Figure 2 illustrates the trade as a share of GDP in Egypt compared to different regions. The Egyptian trade openness seems to increase with time till year Clearly, the trend of the evolution of trade openness is very similar to the one for the Arab world and the MENA region. Percentages went from 40 percent in 2000 to 60 percent in 2008, which shows the effectiveness of trade policies at that time. Starting from year 2008, the figures have changed. A drop in trade accompanies the Economic crisis in 2008 to reach again the 40 percent in Similar trend have also been observed for aggregate exports for which trade have decreased from 30 percent in 2008 to 15 percent in Figure 2: Trade (% of GDP) Egypt MENA Arab World Middle income Source: World Development Indicators Merchandise trade measures the level, year-over-year changes in total trades in goods, exports and imports. For the case of Egypt, as shown in Figure 3, merchandise trade as a share of GDP is reported to be continuously increasing during the last decade. It has more than doubled between years 2000 and 2008; which goes in line with the result shown in Figure 2.

6 Figure 3: Merchandise trade (% of GDP) Egypt MENA Arab World Middle income Source: World Development Indicators As it is shown in Figure 4, Egypt s performance at the manufacturing sector level is better than the one of the MENA region and Arab countries. Moreover, starting 2007, the share of manufactures in merchandise exports has remarkably increased to reach 50 percent in 2009 compared to only 10 percent for the Arab countries and 20 percent for the MENA region. Yet, the Egyptian economy is still lacking behind the middle income countries. Figure 4: Manufactures exports (% of merchandise exports) Egypt MENA Arab World Middle income Source: World Development Indicators

7 Egypt s exports are moderately diversified. Figure 5 shows that almost half of the Egyptian exports are concentrated in fuel, mineral and oil products. Yet, despite a large share in Egyptian exports, proceeds of fuel, mineral oils and products rose by 2% between FY06 and FY10. In the meantime, Egypt managed to diversify its non-oil exports that scaled up due to the increase in exports of raw materials (up by 90.4% over the same period), finished (up by 94.7%) and semi-finished products (up by 36.7%). Clearly, the increase in non-oil exports contributed to the development of the industrial sector that expanded and increased its labor demand for blue-collar workers that are more abundant in Egypt. Figure 5: Proceeds of Merchandise Exports by Degree of Processing 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FY06 FY07 FY08 FY09 FY10 Source: The Central Bank of Egypt Others Finished goods Semi-finished goods Raw Materials Fuel, Mineral Oils & Products The imports structure is a little bit different since fuel and oil products do not represent more than 10% of Egyptian imports. The bulk of imports is concentrated in the categories of raw materials, investment and intermediate goods representing altogether two thirds of imports. Yet, the evolution of imports points out another important fact. As it is shown in Figure 6, both consumer non-durable and durable goods have tripled between FY06 and FY10, especially those coming from China.

8 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Figure 6: Payments for Merchandise Imports by Degree of Use Others Consumer Non-Durable Goods Consumer Durable Goods Investment Goods Intermediate Goods Raw Materials Fuel, Mineral Oils & Products FY06 FY07 FY08 FY09 FY10 Source: The Central Bank of Egypt The geographical distribution of both exports and imports is relatively the same. The European Union is the major trade partner with Egypt, accounting for, on average, 35% of the total trade. The key exports to the EU were crude oil and products, cast iron, cotton textiles, cement, iron and steel products, pharmaceuticals, and aluminum products. The main imports from the EU were crude oil and products, iron and steel products, organic and inorganic chemicals, pharmaceuticals, and electric appliances for telephones and telegraphs. Before the financial crisis, the USA was occupying the second rank for both exports and imports with an average share equivalent to 33% and 22% respectively. The USA chiefly imports from Egypt crude oil and products, cement, and iron and steel products. 100% Figure 7: Geographical Distribution of Exports 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FY02 FY03 FY04 FY05 FY06 FY07 FY08 FY09 FY10 Others Australia Africa Asia Arab USA Russian Federation Other European EU

9 Source: The Central Bank of Egypt Imports from the USA were mainly crude oil and products, iron and steel products, wheat, and maize. Note that starting FY09, the share of imports from Asia outweighed USA s one and came to the second rank after the EU with a 20% share of total imports while the share of American imports declined to reach 11% in FY10 (Figures 7 and 8). Moreover, it is important to note that while Egypt s trade with Asia has been multiplied by 5 between FY02 and FY10, the one with USA has been multiplied by just 1.7 over the same period. The main imports from Asian countries are parts and accessories of cars, animal and vegetable fats, cars, ready-made clothes, and iron and steel products. Finally, the Arab share in Egypt s trade is quite modest since it does not exceed 10% of total trade with crude oil and products, iron and steel products, cast iron, cement, and rice as the main Egyptian exports and crude oil and products, organic and inorganic chemicals, and cars as the main Egyptian imports. Figure 8: Geographical Distribution of Imports 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FY02 FY03 FY04 FY05 FY06 FY07 FY08 FY09 FY10 Others Australia Africa Asia Arab USA Russian Federation Other European EU Source: The Central Bank of Egypt 2.2. Firm-level export dynamics Turning our attention to firms dynamics in the trade market, as it is clearly illustrated in Figure 9, the number of new firms has been divided by 64 between 2010 and 2012 as a result of political uprisings in Egypt. Extensive margin has almost disappeared to reach the level of only 4 new firms in The total number of exporters has also been reduced by 31 percent. Despite the fact that new as well as total exporters have seen the value of their exports improving between 2010 and 2011, this value has dropped from

10 1,661 in 2011 to become 157 in This deterioration in the status of exporters is an expected result of the instable political environment that leads to a disincentive factor for new investors and entrepreneurs to start-up. Figure 9: New Exporters (by year) Source: Ministry of Industry and Foreign Trade, General Organization for Exports & Imports Control (GOEIC), Monthly Digest (2012). Note: A new exporter is defined as the one who did not export for the last three years, provided that the value of annual exports is a million pounds or more. Figure 10 displays the relationship between the number of exporter and the size of the importer that captures the demand at destination. It is worthy to note that there is a positive correlation between both showing that the higher the GDP of the importer, the higher the demand and consequently the higher the number of new exporters. Therefore, the extensive margin of exports is likely to be positively affected by the demand at the destination.

11 Figure 10: Number and mean size of exporters Ln(GDP of Importer) lnumfirm Fitted values Source: Constructed by the authors using the customs data. Table 1: An Overview of Egyptian Exporting Firms Indicator Egypt MENA World Percent of firms exporting directly or indirectly (at least 1% of sales) Percent of firms exporting directly (at least 1% of sales) Proportion of total sales that are domestic sales (%) Proportion of total sales that are exported directly (%) Proportion of total sales that are exported indirectly (%) Days to clear imports from customs* Percent of firms using material inputs and/or supplies of foreign origin* Proportion of total inputs that are of domestic origin (%)* Proportion of total inputs that are of foreign origin (%)* Days of inventory of main input* Percent of firms identifying customs and trade regulations as a major constraint Source: Enterprise Surveys ( The World Bank

12 3. How Red Tape Barriers Do Matter in Egypt? Despite these liberalization efforts, other impediments to trade still exist, especially administrative barriers. Indeed, Egypt has experienced several waves of trade liberalization but its exports and imports are still hindered by either non-tariff measures or implicit barriers to trade. For trade liberalization, the maximum tariff rate has decreased from 110% at the end of the 1980s to reach 40% in the end of 1990's. In 2004, the government of Egypt launched the second wave of liberalization. Its objectives were twofold: first, to reduce tariffs and rationalize the tariff structure; and second, to reduce the number of products subject to non-tariff barriers. The number of tariff bands was narrowed from 27 tariff brackets to 6, tariff dispersion measured by standard deviation declined from 16.1 in 2000 to 12.7 in 2004 and tariff lines were reduced from 8,000 to 6,000. Both nominal and effective protections have declined in the manufacturing sector from 21.3% to 12.1% and from 23.3% to 14% respectively after the 2004 reform. All those measures should in turn simplify procedures, minimize tariff evasion, and remove possibilities of discretion and corruption (Zaki, 2011). Therefore, the increase in exports and imports can be attributable to these trade reforms. Yet, some sectors, such as the food and tobacco sectors, remain highly protected, due to tariff escalation and non-tariff barriers on the trade side, and due to energy subsidies on the input side. The effective rate of protection (ERP) has decreased from 85.6 percent in 1999 to 45 percent in 2009 for private business and from percent to 37 percent for public enterprises over the same period. In addition, they argued that the dispersion of effective rate of protection fell between 1999 and 2009 from 192 to 57 percent, but it remains higher than the low dispersion of nominal tariffs due to first tariffs and output subsidies and second to energy subsidies. Despite a significant liberalization of the manufacturing sector, the primary sector remains relatively protected given the fact that in 2009, its simple average of MFN tariffs is 41% while the manufacturing s one is 9%. Finally, the difference between applied and weighted tariff rates is much larger for the primary sector (37.5% and 6% respectively) than for manufacturing (9.3% and 9.12% respectively). This is due to the fact that some products in the primary sector are subject to high tariffs (such as tobacco and alcohol) whereas their weights in international trade are significantly low.

13 Total Primary Manufacturing Table 2: Tariff Rate by Sector, Applied simple Applied weighted MFN simple MFN weighted Applied simple Applied weighted MFN simple MFN weighted Applied simple Applied weighted MFN simple MFN weighted Source: World Development Indicators. In Egypt, red tape procedures for exports and imports remain high and costly. In 2010, the former request 12 days costing U.S.$ 613 and the latter 12 days adding some U.S.$ 698 to the value of imported goods. Figure 12 shows that the number of documents and days to export and to import is higher than MENA s and OECD s averages. Moreover, Table 3 points out that the documents preparation is the most time-consuming procedure since it requires 60 percent of the total duration. Consequently, Egypt still has a long way to reach better rankings in the ease of doing business or best practice countries in trade facilitation aspects. Table 3: Cost and Duration of Import Procedures Nature of Import Procedures Duration (days) US$ Cost Documents preparation Customs clearance and technical control 1 90 Ports and terminal handling Inland transportation and handling Totals Source: Doing Business Dataset (2013).

14 Figure 11: Trading Across Borders: Egypt vs. MENA vs. OECD Documents to export (number) Documents to import (number) Time to export (days) Time to import (days) Egypt MENA OECD Source: Constructed by the authors using the Doing Business Dataset. In addition, according to the Enabling Trade Index issued by the World Economic Forum (2012), Egypt has been ranked a low 90 th amongst 132 countries for the ease of getting goods across the border. Despite importing goods is not costly, importers raise concerns about the efficiency of customs and other border agencies pointing out to the fact that bureaucracy and transaction length are significant impediments to trade. Its score was 3.78 (the first country is Singapore with a score of 6.14 and the last is Chad with some 2.63). Egypt is not well positioned neither for efficiency of customs administration (ranked 80 th ) nor for transparency of border administration (94 st ). Efficiency of exports and imports is located in a middle position (55 th ). All these facts raise some worries about the efficiency of trade procedures in Egypt. For this reason, policymakers should focus on such barriers to boost foreign trade since its customs administration remains inefficient and corruption-ridden. 4. Methodology The methodology used in this paper draws on the pioneering work of Tinbergen (1962) and Anderson (1979): the gravity model, which became nowadays an essential tool in the empirics of international trade to assess the determinants of trade in goods and services. The gravity model has undergone significant theoretical and empirical improvements over the years (Mac Callum 1995; Feenstra et al. 2001; Feenstra 2002; Anderson and van Wincoop 2003; Evenett and Keller 2002; Santos Silva and Tenreyro 2006), enforcing its theoretical base and thus narrowing the gap between theoretical and empirical findings.

15 Our dependant variable is the value of trade between firm i in Egypt and country j at year t (X ijt ). Our explanatory variables are GDP of Egypt and GDP of partner j, several variables measuring transaction costs that include transport costs measured by the bilateral distance between Egypt and its partner j (d ij ), some dummies capturing whether one country was a colony of the other at some point in time, whether the two have been colonized by a same third country (Comcol ij ) or whether the two countries share a common border (Conti ij ) or share common language (Lang ij ). To controle for other trade policy variables, we introduce the average applied tarrif in the manufacturing sector (Tar j ). We also controle for technological development using the number of secure servers (Internet ijt ): Ln(X ijt )= β 0 + β 1 ln(gdp EGY,t )+ β 2 ln(gdp j,t )+ β 3 ln(d ij )+ β4 Col ij + β 5 Comcol ij + β 6 Conti ij + β 7 Lang ij + β 8 ln(red Tape ijt ) + β 9 ln(tar jt ) + β 10 ln(internet ijt ) + ε ijt (1) where є ijt is the discrepancy term. Similar regressions are run using different dependent variables. The latter capture both the intensive margin (the value of total exports per firm and the average exports per destination) and the extensive margin (the probability of exporting to more than one destination, the number of firms per destination and the number of products per destination). 5. Data We compile our gravity-type variables from different sources. The Gross Domestic Product (GDP) for each country comes from the World Development Indicators database online (2011) that provides GDP in constant 2000 USD. Other classic gravitational variables, for instance contiguity, common language, distance, common colonizer, etc. come from the Centre des Etudes Prospectives et d Information Internationales (CEPII) Distance database (available on Red tape variables (the logarithm of the time to export (to import), the number of documents to export (to import) and the number of procedures to start a business for the exporter (the importer)) are taken from the Doing Business database constructed by the World Bank. are taken also from Doing Business. The internet widespread is taken from the World Development Indicators (WDI). Trade data come from the General Organization for Export and Import Control (GOEIC), the Ministry of Industry and Foreign Trade in Egypt from 2006 to 2010 (at the HS4 level). This Dollar figures for GDP are converted from domestic currencies using 2000 official exchange rates.

16 dataset has four dimensions: exporting firm, year, destination and product for two variables which are value and quantity of exports. It is worthy to note that the World Bank has developed also a series of measures classified under different categories reflecting basic characteristics of the export base in each country (size of the exporting sector, exporter size and exporter growth rates), concentration/diversification (Herfindahl index, share of top exporters, number of products and destinations per exporter), firm, product and market dynamics (entry, exit and survival rates) and unit prices (per exporter, product, market). The measures are available at different levels of aggregation, including: a) country-year, b) country-year-product, and c) country-year-destination. However, one drawback of this data is that we cannot explore the link between export behavior and firms performance measures. Such analysis may be conducted if the exporter-level transaction data can be merged with industrial census data including key firm characteristics such as employment, profits, gross output per worker and wages. 6. Empirical Results In order to examine the effect of red tape costs on exports performance, we run several regressions with different techniques and different dependent variables. First, to capture the intensive margin of exports, we use the value of total exports per firm and the average exports per destination. Second, for the extensive margin, we use the probability of exporting to more than one destination, the number of firms per destination and the number of products per destination. First, as it is shown in Table 5, our gravity model is doing well since both the exporter and the importer GDP, in most of the regressions, have the expected positive sign and are statistically significant. As expected, bilateral distance has a negative impact on trade in goods. Moreover, our policy variable, which the applied tariff, has also a statistically negative effect on bilateral exports, especially in the fixed effect regression. While the variable of colonial links seems to increase the Egyptian exports, common language and borders, surprisingly, do have a negative impact on bilateral exports. This result is due to the fact that most of the countries that share common borders with Egypt have political problems that prevent them from trading in a significant way with Egypt. Finally, it is worthy to mention that in the fixed effect regressions, all the variables that are time invariant (distance, common language, colony and contiguity) are automatically dropped since they are captured by the fixed effects. As per our variables of interest, Table 5 shows that both time to export and time to import have a significant negative impact on bilateral exports (in the fixed effect regressions). Moreover, the effect of time to export is greater than the one faced by Egyptian firms in the destination country. Thus, the exporting country benefits more than the importing one

17 from trade facilitation. In other words a country gains more from its own trade facilitation than from its partner one. This conclusion was proved by Wilson et al. (2004) where 75% of the benefits of any measure are devoted to the exporter while the remaining 25% to the importer. For example, thanks to the improvement of ports efficiency, trade flows increase by 2.2% for the exporter and 0.6% for the importer. Similarly, by improving the infrastructure of the service sector, the exporter's trade flows rise by 3%, while those of the importer by 1%. However, a multilateral facilitation may amplify such gains. Table 5: The incidence of RTC on the Value of Exports (Intensive Margin) FE RE FE RE ln(eg gdp) *** *** (0.431) (0.343) (0.419) (0.408) ln(importer gdp) 1.152*** 0.220*** 1.318*** 0.225*** (0.281) (0.0410) (0.314) (0.0464) Contiguity *** *** (0.220) (0.228) Colony 0.778*** 0.799*** (0.115) (0.118) common language ** *** (0.135) (0.132) ln(distance) * ** (0.0790) (0.0790) ln(internet) *** *** (0.0499) (0.0302) (0.0567) (0.0385) ln(eg time to X) ** *** * *** (0.106) (0.0955) (0.125) (0.131) ln(time to import) ** 0.149* * (0.0826) (0.0865) (0.0927) (0.0879) ln(tariff) ** 2.253* (1.451) (1.243) Constant ** *** ** *** (13.65) (9.249) (14.19) (10.82) Observations R-squared Number of id Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

18 It is worthwhile to note that a Hausman test between fixed-effect and random-effect was done. We did not fail to reject the null hypothesis that there is a systematic difference between both models, thus preferring the fixed-effect one. As per the extensive margin (see Table 6), the results are different since both contiguity and common language are likely to increase the probability of becoming an exporter to a certain destination. Surprisingly, the effect of time to export is positive. By contrast, the number of days faced by the exporter at the country of destination has a statistically significant impact on the firm s exports. Table 6: The incidence of RTC on the Probability of Exporting (Extensive Margin) FE RE FE RE ln(eg gdp) *** *** *** *** (0.218) (0.178) (0.269) (0.220) ln(importer gdp) 1.440*** *** 1.461*** *** (0.110) ( ) (0.141) ( ) Contiguity 0.117** 0.144** (0.0559) (0.0608) Colony (0.0433) (0.0448) Common language 0.127*** 0.120*** (0.0177) (0.0202) ln(distance) (0.0101) (0.0119) ln(internet) *** *** * (0.0201) ( ) (0.0276) ( ) ln(eg time to X) 0.287*** 0.307*** 0.336*** 0.133** (0.0580) (0.0569) (0.0711) (0.0671) ln(time to import) *** *** ** (0.0356) (0.0146) (0.0413) (0.0173) ln(tariff) *** *** (0.923) (0.218) Constant 14.80*** 33.76*** (4.758) (5.862) lnsig2u *** *** (0.0188) (0.0207) Observations Number of id Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

19 The incidence of red tape costs on average exports that measure the intensive margin as well is negative as well. Indeed, Table 7 confirms the results we obtained in Table 5 since the time to export seems to have a statistically negative impact on bilateral exports. By contrast, the one faced at the destination is insignificant. Table 7: The incidence of RTC on Average Exports by Destination FE RE ln(eg gdp) (1.099) (1.022) ln(importer gdp) *** (0.610) (0.0594) Contiguity (0.923) Colony (0.889) common language *** (0.199) ln(distance) *** (0.0899) ln(internet) * (0.0993) (0.0511) ln(eg time to X) *** *** (0.355) (0.351) ln(time to import) (0.150) (0.113) ln(tariff) (3.444) (1.435) Constant (34.63) (27.43) Observations R-squared Number of ISO id Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

20 Results in Table 8, 9 and 10 present different measures of the extensive margin. The main message is that effect of red tape cost on the extensive margin of exports is not robust. In fact, while the effect of time to export on the number of products per destination is positive and slightly significant, the one to import is not significant. Moreover, a similar pattern is observed for the number of firms per destination. Finally, the effect on red tape costs on the number of products per firm and per destination is not significant at all. Table 8: The incidence of RTC on the Number of Products per Destination FE RE FE RE ln(eg gdp) (0.493) (0.469) (0.544) (0.523) ln(importer gdp) 1.094*** 0.677*** 1.314*** 0.657*** (0.239) (0.0537) (0.267) (0.0573) Contiguity 3.341** 3.457*** (1.369) (1.329) Colony *** *** (0.792) (0.765) common language 1.814*** 1.906*** (0.304) (0.303) ln(distance) *** *** (0.137) (0.135) ln(internet) * (0.0375) (0.0329) (0.0452) (0.0382) ln(eg time to X) 0.261* 0.232* 0.277* (0.140) (0.139) (0.153) (0.153) ln(time to import) (0.0796) (0.0717) (0.0856) (0.0780) ln(tariff) (1.616) (1.319) Constant *** * ** (13.74) (12.81) (15.84) (14.28) Observations R-squared Number of ISO_id Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

21 Table 9: The incidence of RTC on the Number of Firms per Destination FE RE FE RE ln(eg gdp) 0.974* 0.931** (0.504) (0.444) (0.511) (0.483) ln(importer gdp) 1.074*** 0.636*** 1.277*** 0.597*** (0.248) (0.0497) (0.310) (0.0490) Contiguity 2.777* 2.956** (1.466) (1.507) Colony (1.445) (1.478) common language 1.567*** 1.697*** (0.195) (0.203) ln(distance) *** *** (0.105) (0.107) ln(internet) * (0.0490) (0.0397) (0.0510) (0.0445) ln(eg time to X) 0.326*** 0.293** 0.328*** 0.250** (0.124) (0.123) (0.125) (0.127) ln(time to import) * (0.0721) (0.0695) (0.0774) (0.0782) ln(tariff) (1.994) (1.665) Constant *** ** ** (13.88) (12.34) (15.46) (13.33) Observations R-squared Number of ISO id Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

22 Table 10: The incidence of RTC on the Number of Products per Firms and Destination Eq1 Eq2 ln(importer gdp) *** *** ( ) ( ) Contiguity 0.233*** 0.251*** (0.0205) (0.0225) Colony ** *** (0.0158) (0.0160) common language 0.182*** 0.193*** ( ) ( ) ln(distance) *** *** ( ) ( ) ln(internet) *** *** ( ) ( ) ln(time to import) ( ) ( ) ln(tariff) *** (0.0838) Year dummies YES YES Constant *** * (0.0584) (0.0654) Observations R-squared Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

23 Table 11 presents the results of the quantile regressions. In this set of regressions, we try to examine the differential impact of red tape costs on the value of exports. The administrative barriers faced by Egyptian firms the country of destination have a negative and significant effect on firms exports. It is worthy to mention that both of the GDP of exporter and the time to export at the country of origin are dropped since they are captured by the year dummies. Table 11: Results of Quantile Regressions showing the incidence of RTC on the Value of Exports 90th 75th 50th 25th 10th ln(importer gdp) 0.307*** 0.279*** 0.207*** 0.257*** 0.254*** (0.0211) (0.0176) (0.0157) (0.0208) (0.0274) Contiguity *** *** *** *** *** (0.151) (0.126) (0.112) (0.149) (0.197) Colony 0.900*** 0.903*** 0.833*** 0.971*** 0.869*** (0.108) (0.0898) (0.0798) (0.106) (0.140) common language *** *** *** *** *** (0.0510) (0.0425) (0.0378) (0.0503) (0.0662) ln(distance) *** *** *** *** *** (0.0300) (0.0250) (0.0223) (0.0296) (0.0390) ln(internet) *** *** *** *** *** (0.0200) (0.0166) (0.0148) (0.0197) (0.0259) ln(time to import) *** ** 0.158*** 0.320*** 0.312*** (0.0453) (0.0377) (0.0335) (0.0446) (0.0587) ln(tariff) 3.510*** 3.219*** 2.226*** 2.592*** (0.567) (0.473) (0.420) (0.560) (0.736) Year dummies YES YES YES YES YES Constant 9.353*** 7.866*** 7.071*** 4.539*** 2.254*** (0.446) (0.372) (0.331) (0.440) (0.579) Observations Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 In a nutshell, the findings of this paper show that red tape barriers negatively affect Egyptian firms. This effect seems to be more robust for the intensive margin (the value of exports) rather than the extensive one (number of products and firms per destination). Yet, the impact of administrative barriers to trade on the number of destinations by firm is likely to be significantly negative.

24 7. Conclusion and Policy Implications This paper tackles the impact of red tape barriers on firms exports. The topic of this paper is crucial in international trade for three main reasons: first, trade barriers- as argued by the WTO- are highly correlated to lengthy, bureaucratic and time consuming trade procedures that do negatively affect firms exports. Second, these barriers are significantly highly persistent and costly in developing countries such as Egypt. In addition, they represent a deadweight loss as they do not generate any rent or revenue. In the present study, we estimate a gravity model using Egyptian firm-level data to examine the impact of these barriers on firms exports. For administrative barriers, we use the Doing Business dataset that has been developed by the World Bank. The preliminary findings show that red tape barriers negatively affect Egyptian firms. This effect seems to be more robust for the intensive margin (the value of exports) rather than the extensive one (number of products and firms per destination). Yet, the impact of administrative barriers to trade on the number of destinations by firm is likely to be significantly negative. Moreover, the project findings based on firm-level data have several implications for policy-makers in the region concerned with promoting private sector development, exports and job creation. Policies that lower trade costs and favor access to export markets can trigger a selection process whereby the most productive firms substitute the least productive ones within sectors. This would in turn be beneficial for productivity and job creation, since exporters perform better in both areas. References [1] Anderson, J. (1979) A Theoretical Foundation for the Gravity Equation, American Economic Review vol. 69, pages [2] Anderson, J. and van Wincoop, E. (2004) Trade Costs, Journal of Economic Literature vol. 70(1), pages [3] Anderson, J. and van Wincoop, E. (2003) Gravity with Gravitas: A Solution to the Border Puzzle, American Economic Review vol. 93(1), pages [4] Djankov, S., Freund, C. and Pham, C. (2006) Trading on Time, Policy Research Working Paper Series No. 3909, World Bank, January. [5] Feenstra, R. (2002) Border Effects and the Gravity Equation: Consistent Methods for Estimation, Scottish Journal of Political Economy, vol. 49(5), Novembre.

25 [6] Fontagné, L. and Zignago, S. (2007) A Re-evaluation of the Impact of Regional Agreements on Trade Patterns, September, Integration and Trade, No. 26, January-June. [7] Freund, C. and Weinhold, D. (2000) On the Effect of the Internet on International Trade, International Finance Discussion Papers, No. 693, Board of Governors of the Federal Reserve System. [8] Fujita, M., Krugman, P. and Venables, A. (2000) The Spatial Economy: Cities, Regions, and International Trade, Southern Economic Journal, vol. 67(2), pages , October. [9] Hummels, D. (2001) Time as a Trade Barrier., Department of Economics, Indiana: Purdue University, mimeo. [10] McCallum, J. (1995) National Borders Matter : Canada-U.S. Regional Trade Patterns, The American Economic Review, vol. 83(2), pages [11] O.E.C.D. (2002b) Avantages pour les Entreprises de la Facilitation des Echanges, prepared by T. Matsudaira and Evokia Mosé, Working Party of the Trade Committee, TD/TC/WP(2001) 21/FINAL, August.